This project will investigate DNA base mismatch correction in Diplococcus pneumoniae. Such repair occurs after DNA-mediated genetic transformation and after potentially mutagenic base substitution in DNA replication. In transformation, single-site markers are integrated with different efficiencies because a system, called Hex, recognizes mismatches in the intial heteroduplex product of integration and removes the donor portion with a frequency depending on the mismatched base pair formed. Hex(minus) mutants give only high integration efficiency. The molecular mechanism of the Hex system will be determined in vivo by following the biochemical fate of a purified DNA donor fragment containing the amylomaltase gene. This gene has been extensively analysed genetically. By design of a suitable substrate and use of cell extracts lacking nonspecific nucleases, we hope to demonstrate mismatch repair in vitro. The enzymatic mechanism of the repair process could then be analyzed. Of particular importance is how the system recognizes and excludes the donor contribution to the mismatch. Strains that lack the mismatch repair system show higher spontaneous mutation rates. The system apparently repairs DNA replication errors by recognizing base mismatches and removing the wrong base in the nascent strand. By analysing reversion rates of mutations in the amylomaltase gene, the mismatch specificities of mutational correction and integration efficiency will be compared. Mismatch repair, as indicated by the occurrence of gene conversions, is widespread among living cells. The proficiency of this system could be a major factor in determining spontaneous mutation rates. The impact of DNA methylation on spontaneous mutation rates will also be ascertained. If ageing is due to the accumulation of mutations, mismatch repair could be important for longevity and for the prevention of age-related diseases such as cancer and heart disease. An in vitro assay will make the study of mismatch repair accessible also in human cells.